RES EARCH

◥ cation are less likely to be mentally ill, richer

REVIEW SUMMARY
ECONOMICS
Poverty, depression, and anxiety: Causal evidence
and mechanisms
Matthew Ridley, Gautam Rao, Frank Schilbach*, Vikram Patel
BACKGROUND: Depression and anxiety disor-
ders are together responsible for 8% of years
lived with disability globally. Contrary to
widely held preconceptions, these are not
diseases of affluence. Within a given location,
those with the lowest incomes are typically
1.5 to 3 times more likely than the rich to
experience depression or anxiety.
that randomized interventions to treat men-
tal illnesses increase days worked. Depression
and anxiety directly affect the way people
think, by capturing their attention and dis-
torting their memory. Such effects are likely
to influence economic preferences and beliefs
and thus distort important economic deci-
sions made by individuals, such as how much
countries do not have lower rates of mental
illness. Thus, aggregate economic growth
alone is unlikely to reduce mental illness. Cli-
mate change is likely to worsen mental health,
both directly through the effect of higher tem-
peratures on mood and through reductions
in agricultural yields because of changes in
rainfall and water supply, more frequent
weather-related disasters, and an increased
likelihood of violent conflict. Technological
change and globalization create large over-
all economic gains but also concentrated
groups of losers whose mental health may
be compromised. The spread of social me-
dia and associated technologies may also
be harming mental health, especially among
adolescents.
Policy action on mental health is vital, as is
interdisciplinary research on the mechanisms

Downloaded from http://science.sciencemag.org/ on December 10, 2020

to work, invest, and consume. Reduced con- that link poverty and mental illness. Recently
ADVANCES: Recent research has established a centration and greater fatigue reduce work developed approaches to psychotherapy, de-
bidirectional causal relationship between pov- productivity, and the social stigma of men- livered through nonspecialist providers, pro-
erty and mental illness. Researchers vide a scalable and effective approach
have begun to isolate the underlying to improving mental health in low-
mechanisms, which can guide effec- Poverty income countries. Given the associ-
tive policies to protect the mental ated economic benefits of improved
health of those living in poverty. mental health, such interventions
We now know that loss of income should be a part of the antipoverty
causes mental illness. Negative in- Productivity, labor supply Worry toolkit alongside more traditional
come shocks, such as bad harvests Preferences and beliefs Physical health economic interventions. Understand-
due to poor rainfall or job losses due ing the most effective combination
Economic decision-making Early-life conditions
to factory closures, worsen mental of economic and psychological sup-
health. Conversely, cash transfers and Women’s empowerment Violence and crime port in different populations is an
broader antipoverty programs reduce Childhood development Social status important next step. A priority for re-
depression and anxiety in randomized search is testing for a mental health–
trials. Multiple mechanisms mediate based “poverty trap.” If such poverty
this causal chain. Poverty is associ- traps exist, then powerful one-time
ated with volatile income and ex- interventions will have large long-
penditures. The resulting worries and Mood and anxiety run effects as gains in mental health
uncertainty can worsen mental health. disorders and economic outcomes reinforce
Providing health, employment, or one another. Evaluations of economic
weather insurance, or other ways of The causal relationship between poverty and common mental interventions should routinely mea-
smoothing shocks, may thus lower illnesses. This schematic shows the principal mechanisms we identify, sure mental health, and long-run
depression and anxiety. Living in on the basis of theory and empirical evidence, through which poverty evaluations of mental health inter-
inadequate housing in low-income and depressive and anxiety disorders interact. ventions should measure potential
neighborhoods, the poor are also impacts on poverty and other key
more exposed to environmental stresses such tal illness may further worsen labor-market economic outcomes. The causal relationship
as pollution, temperature extremes, and chal- outcomes. Mental illness appears to increase between poverty and mental health is even
lenging sleep environments, which can cause the likelihood of catastrophic health expendi- more pertinent given the ongoing pandemic,
mental illness. Early-life conditions—poverty tures for individuals through its comorbidity which has disproportionately affected the
experienced in childhood and in utero— with chronic illnesses such as diabetes and poor and may have lasting impacts on their
increase the likelihood of poor nutrition and heart disease. Mental illness may also hinder economic and mental well-being. A massive
other stressors, resulting in impaired cog- education and skill acquisition among youth investment in mental health was already
nitive development and adult mental ill- and exacerbate gender inequalities through long overdue. It has now become critically
ness. This makes a strong case for providing
financial support to pregnant women and
its disproportionate prevalence among women.
Parental mental illness can also influence
urgent.
▪
caregivers of young children. Poverty is also children’s cognitive development and educa-
associated with worse physical health; greater tional attainment, transmitting mental illness The list of author affiliations is available in the full article online.
and poverty across generations. *Corresponding author. Email: fschilb@mit.edu
exposure to trauma, violence, and crime; and
Cite this article as M. Ridley et al., Science 370, eaay0214
lower social status, each of which may af- (2020). DOI: 10.1126/science.aay0214
fect mental health. OUTLOOK: The burden of mental illness is
Mental illness in turn worsens econo- likely to increase in the coming decades. READ THE FULL ARTICLE AT
mic outcomes for individuals. Studies show Although richer individuals within a given lo- https://doi.org/10.1126/science.aay0214

Ridley et al., Science 370, 1289 (2020) 11 December 2020 1 of 1

RES EARCH

◥ What are the causal links between poverty

REVIEW and mental illness? Can economic policies im-
prove psychological well-being? Can psycho-
ECONOMICS logical interventions reduce poverty? Any
attempt to understand this relationship must
Poverty, depression, and anxiety: Causal evidence acknowledge the complexity and multidimen-
sional nature of both mental health and
and mechanisms poverty. Mental health in the broadest possi-
ble sense has been defined as “a state of well-
Matthew Ridley1, Gautam Rao2, Frank Schilbach1*, Vikram Patel3,4 being in which the individual realizes his or
her own abilities, can cope with the normal
Why are people who live in poverty disproportionately affected by mental illness? We review the stresses of life, can work productively and
interdisciplinary evidence of the bidirectional causal relationship between poverty and common fruitfully, and is able to make a contribution
mental illnesses—depression and anxiety—and the underlying mechanisms. Research shows that to his or her community” (18). This definition
mental illness reduces employment and therefore income, and that psychological interventions includes both happiness or life satisfaction,
generate economic gains. Similarly, negative economic shocks cause mental illness, and antipoverty which also correlate positively with income
programs such as cash transfers improve mental health. A crucial step toward the design of (19), and symptoms associated with anxiety
effective policies is to better understand the mechanisms underlying these causal effects. and mood disorders, such as depression. The
two are clearly related; depression and anxiety

D
are strong determinants of happiness (20),

Downloaded from http://science.sciencemag.org/ on December 10, 2020

epression and anxiety are the most com-
mon mental illnesses: 3 to 4% of the
world’s population suffers from each
at any given time, and they are together
responsible for 8% of years lived with
disability globally (1). Contrary to widely held
preconceptions from the 20th century, these
are not “diseases of affluence” (2, 3). Within a
given location, those living in poverty are at
least as likely to suffer as the rich. By some
measures, the poor are substantially more
likely than the affluent to experience mental
ill-health. Rates of depression, anxiety, and
suicide correlate negatively with income (4–7)
and employment (5, 8). Those with the lowest
incomes in a community suffer 1.5 to 3 times
more frequently from depression, anxiety, and
other common mental illnesses than those
with the highest incomes (5). For example,
in India, 3.4% of those in the lowest income
quintile experience depression at any given
time, compared with 1.9% of those in the
highest quintile (Fig. 1).
In this Review, we explore the evidence for
the bidirectional causal relationship between
poverty and mental health and its underlying
years lived with disability globally are known
to be caused by depression, anxiety, and other
mental illnesses (1). Low- and middle-income
countries spend an even smaller share of al-
ready small health budgets on mental health
(Fig. 2). Despite the existence of cost-effective
treatments, such low investments in mental
health have contributed to treatment gaps of
more than 80% globally for common mental
illnesses, which is much larger than for major
physical health conditions (11–14). Mental and
physical health are tightly connected: When
mental health problems coexist with physical
health problems, health outcomes, disability,
and costs tend to be much worse (15–17). How-
ever, unlike most physical health conditions,
mental disorders may directly distort economic
decision-making in ways that perpetuate pov-
erty, by directly affecting cognitive function,
preferences, and beliefs.
Current prevalence
4
Prevalence of depression (%)
and ultimately, mental health and even men-
tal illnesses such as depression and anxiety
exist along a continuum.
We focus on the causal evidence that links
poverty with depressive and anxiety disorders,
the most common mental illnesses, which we
refer to here using the more general terms
“mental health” and “mental illness.” Defini-
tions of these illnesses and a brief primer on
their measurement are provided in Box 1.
Although other more serious mental illnesses,
such as schizophrenia, are also correlated with
poverty, and may have powerful effects on
economic outcomes, we do not discuss them
here (21).
Like mental health, poverty is multidimen-
sional. We examine causal links between men-
tal illness and important economic dimensions
of poverty, particularly income and unemploy-
ment. We also touch on other dimensions
±1 SEM

mechanisms. Poverty is also correlated with

poor physical health (9, 10), but the relation- 3
ship between mental illness and poverty has
been overlooked and is worth emphasizing.
Mental health has historically not been con- 2
sidered a priority by economists and policy-
makers, and until recently, mental health
care had not been evaluated as an antipoverty 1
tool. Mental health services are underresourced
relative to physical health systems. On aver-
age, countries spend 1.7% of their health bud-
0
gets on mental health, even though 14% of Lowest Second Third Fourth Highest
1
Department of Economics, Massachusetts Institute of Income quintile
Technology, Cambridge, MA 02139, USA. 2Department of
Economics, Harvard University, Cambridge, MA 02138, USA. Fig. 1. Prevalence of depression by income quintile in India. The average percentage of people in each
3
Harvard Medical School, Harvard University, Boston, MA income quintile in India who have had depression within the past 2 weeks (“current” prevalence). Error
02115, USA. 4Harvard TH Chan School of Public Health,
Harvard University, Boston, MA 02115, USA. bars show ±1 SEM. These numbers come from (130), an analysis of the Indian National Mental Health
*Corresponding author. Email: fschilb@mit.edu Survey, 2015–2016.

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 1 of 12

RES EARCH | R E V I E W

4 mental illness with consumption appears to be

weaker than that between mental health and
income (22–25). Income is more volatile than

(% of total health expenditure)

Mental health expenditure consumption in the short run. The stronger
3 correlation of mental health with income sug-
gests that mental health may be more affected
by short-run changes to economic status than
2 long-run or permanent changes.
We discuss evidence on poverty-alleviation
programs and mental health treatments ob-
tained from randomized controlled trials (RCTs).
1 These generate variation in individuals’ pov-
erty and mental health status, respectively,
that is entirely by chance and therefore un-
0 correlated with all other shared risk factors.
Low Lower middle Upper middle High Such studies allow us to isolate evidence of
Income category causal relationships. We also discuss studies
of “natural experiments” in which naturally
Fig. 2. Mental health expenditure by country income category. The average percentage of occurring variation in economic circumstan-
overall health budgets spent on mental health across countries in each of the four income ces or mental health is argued to be “as good

Downloaded from http://science.sciencemag.org/ on December 10, 2020

categories used by the World Bank. Percent spent on mental health comes from the authors’ own as random.” Examples range from financial
calculations, using data on overall mental health spending from the WHO Mental Health Atlas 2017 windfalls, such as lottery wins—for which
(www.who.int/mental_health/evidence/atlas), with data on total overall health spending from lottery winners may be thought of as a treat-
the WHO’s Global Health Expenditure Database (https://apps.who.int/nha/database). ment group and lottery losers as a control
group—to weather shocks that affect some
farmers’ incomes more than others’.
Box 1. Definition and measurement of depression and anxiety.
The causal impact of poverty on mental
Depression, by which we refer here to major depressive disorder, is a constellation of symptoms ill-health
that includes changes in psychomotor function, weight loss, oversleeping or undersleeping, decreased Job loss and income declines are drivers of
appetite, fatigue, difficulty concentrating, extreme feelings of guilt or worthlessness, and suicidal poverty and often precede episodes of mental
thoughts. According to the American Psychiatric Association’s Diagnostic and Statistical Manual of illness (26, 27). Evidence from natural exper-
Mental Disorders (DSM-5), diagnosis of depression requires a set of these symptoms to be present iments confirms that this relationship is causal.
over a 2-week period. For example, reduced agricultural output and
Anxiety, by which we refer here to generalized anxiety disorder, is characterized in the DSM-5 by income because of extreme rainfall caused
long-lasting and excessive fear and worries over at least a 6-month period, with three or more of the increased rates of depression and suicide in
following symptoms: restlessness, fatigue, concentration problems, irritability, muscle tension, and rural parts of Indonesia (Fig. 3 and Box 2) (28).
problems with sleep. Other definitions require the presence of at least one physical symptom such as Similarly, job losses because of plant closures
heart palpitations, difficulty breathing, nausea or abdominal distress, dizziness, and numbness. in Austria were associated with higher subse-
Measuring depression and anxiety in large population samples is feasible by using nonspecialist quent antidepressant use and mental health–
surveyors or even through self-administration of questionnaires. Reliable short-form diagnostic tools related hospitalization (29). Areas in the United
can predict professional diagnosis with rates of false positives and false negatives of 10 to 20% and States more exposed to trade liberalization
have been validated in low-income countries (151–154). Widely used tools include the Generalized with China saw reduced income and employ-
Anxiety Disorder 7-item (GAD-7) scale for anxiety, the Patient Health Questionnaire (PHQ-9) for ment for some groups of workers and in-
depression, or the Self-Reporting Questionnaire 20-Item (SRQ-20) scale for any common mental creased mortality through drug overdoses
illness. These scales typically ask respondents how much they experienced symptoms of depression or among those same groups (30). Whether
anxiety (such as sadness, lack of concentration, or poor sleep) in the past few weeks. The PHQ-9 and job loss worsens mental health beyond the
GAD-7 ask one question for each of the symptoms that are used to define major depressive disorder impacts of the associated loss of income is
and generalized anxiety disorder, respectively. In practice, depression and anxiety are correlated, as unclear, but both mechanisms are argued
evidenced by the fact that they share some symptoms. to play a role in the phenomenon of “deaths
The Center for Epidemiologic Studies Depression Scale (CES-D) is a popular measure among studies of despair” (31).Conversely, income or wealth
of the effect of economic interventions or shocks on mental health. Several studies also use custom increases can improve mental health. For ex-
indices of psychological well-being, typically an average of a life satisfaction scale, a “stress index,” and ample, Native American tribes that opened
some measure of worry or anxiety. In practice, such indices often measure several of the same casinos have seen substantial rises in income
symptoms as the PHQ-9 and GAD-7. and reductions in anxiety relative to those that
Some national surveys already include short-form screening tools, such as the UK Longitudinal did not (32). Some studies have shown that
Household Panel Survey and the South Africa National Income Dynamics Study. lottery winners enjoy better mental health
compared with those who win less or play
but do not win (33). However, when fully
of poverty, including a lack of capabilities tionship between mental health and the con- controlling for the number and frequency
resulting from low education and physical sumption of goods and services, which is a of lottery tickets bought, winnings have small
health, as well as relative poverty and asso- more direct economic measure of poverty. The or no impacts on mental health (34, 35).
ciated low social status. Because of a relative existing evidence of this relationship is con- The most compelling causal evidence that
scarcity of studies, we focus less on the rela- tentious, but the cross-sectional correlation of poverty causes mental illness comes from

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 2 of 12

RES EARCH | R E V I E W

Reductions in suicide rates due to cash transfers rams in other settings with similar approaches
have found similar effects (39, 40). Longer-run
1 ***
Reduction in suicides per 100,000
| | effects of such programs, when measured,
appear to be even larger for both economic
0.8 outcomes and mental health. In India, for ex-
ample, an index of psychological well-being
was 0.24 SD higher in the treatment group
0.6 7 years after the completion of a graduation
program (41).
Hence, across a wide range of populations
0.4
and study designs, positive economic shocks
to individuals are shown to improve mental
0.2 health, whereas negative economic shocks
undermine mental health. This robust evi-
0
dence, on the effects of changes in economic
Average effect Effect for districts Effect for districts circumstances, indicates that poverty does
for all districts not in drought in drought cause mental illness. However, with two ex-
Treatment effect of cash transfers ceptions (35, 41), the above studies consider
the consequences of changes in economic
Fig. 3. Cash transfers, suicide rates, and droughts. The estimated effect of the cash transfer roll-out on status for a few years at most. An important

Downloaded from http://science.sciencemag.org/ on December 10, 2020

district suicide rates, for all districts and separately by whether or not they were experiencing a drought question is whether these short-run effects can
(bottom 20% of the rainfall distribution) when the cash transfers reached them. Error bars show ±1 SEM. persist or grow over time. For example, some
Asterisks denote a significant difference between effects: ***P < 0.01. of the causal mechanisms we discuss below
could take decades to manifest. However,
there may be a hedonic adaptation effect in
which mental health eventually adapts to
the change in circumstances, so that even
Box 2. Cash transfers, rainfall shocks, and suicides.
permanent increases in one’s income level
Christian, Hensel, and Roth examined how income shocks affect suicide rates and depression in have a limited long-run effect. Of course, it
Indonesia (28). They examined two natural experiments: the staggered roll-out across subdistricts of a could also be that the positive economic
conditional cash transfer program and annual and spatial variation in rainfall that affects farmers’ shocks are themselves undone by future nega-
incomes. They measured depression using a 10-question CES-D scale that is included in the Indonesian tive shocks, causing mental health to revert
Family Life Survey. They also used the incidence of suicides, as measured by the reports of village to initial levels or even worsening. Ongoing
leaders in census surveys. long-run evaluations of cash-transfer prog-
Subdistricts that received the cash transfer program in the first wave of roll-out saw an 18% drop in rams are expected to provide evidence on
suicides (P < 0.01) relative to those that received it later, even though both sets of districts had similar this question (42).
trends in suicide before the program’s start. Meanwhile, rural subdistricts that experienced excess Mechanisms for poverty causing mental ill-health
rainfall that increased crop yields between census years saw decreases in depression and suicides
relative to subdistricts experiencing drought. The cash transfer had its largest effects on suicide in How does poverty cause mental illness? We
districts undergoing droughts, suggesting that policy can play a role in mitigating the mental health discuss several plausible causal mechanisms
effects of economic shocks (Fig. 3). and the limited evidence for each. The worries
Because suicide was only measured at the subdistrict level in this study, it is not possible to fully and uncertainty that come with living in pov-
disentangle the direct effects of the cash transfers on recipients from spillover effects to others in the erty seem to be an important driver of mental
village. This highlights the need for better routine data collection on mental health outcomes alongside illness, as do the effects of poverty on child-
economic variables. hood development and one’s living environ-
ment. We have more limited causal evidence
for other plausible channels, including wor-
sening of physical health caused by poverty,
RCTs that evaluate antipoverty programs. completion and 0.16 SD lower after almost increased exposure to violence or crime, and
Several studies that evaluated cash transfer 3 years, with larger transfers causing sub- the effects of low relative social status and
and broader antipoverty programs have found stantially larger effects. social isolation. Understanding which of these
substantial positive impacts on mental health, Similarly, multifaceted antipoverty pro- mechanisms are important may have implica-
including over long time horizons, after the grams beyond cash transfers yield mental tions for policy choices. For example, if worries
effects of any initial celebratory reactions health benefits. A recent large-scale random- and uncertainty play a major role, then provid-
among recipients have worn off. A meta- ized evaluation of a “graduation program” ing health and unemployment insurance may
analysis of these studies is provided in Fig. 4. in six countries that provided extremely poor be crucial, whereas if early-life conditions are
For example, cash transfers to Kenyan house- participants with a mix of assets, intensive the key drivers, then cash transfers to parents
holds worth $400 to $1500 at purchasing training, temporary cash support, savings of young children could be the most appropri-
power parity (about 3 to 12 months of house- incentives, and help to access health care ate policy response.
hold income) increased consumption and hap- found increases in consumption and assets
piness while reducing depression, stress, and 3 years later. The program also improved an Worries and uncertainty
worries (36, 37). Scores on a depression scale index of psychological well-being by 0.1 SD, The anticipation of economic shocks, not
were 0.12 standard deviations (SD; closely which was driven by an increase in happiness just their occurrence, may cause mental ill-
related to Cohen’s d) lower 4 months after and a decrease in mental distress (38). Prog- ness. People living in poverty face substantial

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 3 of 12

RES EARCH | R E V I E W

Study Country Outcome Years elapsed since: Intervention cost in:

Program Start Program End $ MER $ PPP
Multifaceted antipoverty programs
Blattman et al. (2019) Ethiopia PWB 5 4 450 1291

Edmonds and Theoharides (2020) Philippines PWB 3 0.75 184 518

Green et al. (2016) Uganda APAI−R 1.3 − 874 2150

Janzen et al. (2018) Nepal PWB 1.5 − 137 478

Banerjee et al. (2015) Multiple PWB 3 1 1467 3717

Bandiera et al. (2017) Bangladesh PWB 4 2.5 302 1120

Ismayilova et al. (2018) Burkina Faso CES−D 2 1 100 279

Glass et al. (2017) DRC HSCL 1.5 − 40 66

Banerjee et al. (2016) India PWB 7 5.5 357 1257

Bedoya et al. (2019) Afghanistan PWB 2 1 1688 6198

Cash transfers
Heath et al. (2020) Mali PSS 3 − 649 1468

Fernald and Hidrobo (2011) Ecuador CES−D 1.4 − 255 676

Hjelm et al. (2017a) Zambia PSS 3 − 396 816

Macours and Vakis (2009) Nicaragua CES−D 0.75 − 260 758

Kilburn et al. (2019) South Africa CES−D 3 − 540 870

Downloaded from http://science.sciencemag.org/ on December 10, 2020

Baird et al. (2019) Malawi GHQ−12 4.3 2.3 180 440

Blattman et al. (2017) Liberia APAI−R 1 0.8 341 716

Haushofer et al. (2020) Kenya PWB 1 1 150 338

Blattman, Fiala and Martinez (2019) Uganda PWB 9 9 382 1175

Shapiro (2019) Kenya PWB 0.75 0.5 122 314

Hjelm et al. (2017b) Zambia PSS 3 − 432 891

Egger et al. (2019) Kenya PWB 1.5 1.5 1000 1871

Baird et al. (2013) Malawi GHQ−12 2.3 0.3 180 440

Kilburn et al. (2016) Kenya CES−D 4 − 960 2370

Aguilar (2012) Mexico SWB 5 2 900 1651

Haushofer and Shapiro (2018) Kenya PWB 3.4 3 521 709

Haushofer, Mudida and Shapiro (2020) Kenya PWB 1 1 534 1184

Angeles et al. (2019) Malawi CES−D 2 − 156 517

Multifaceted antipoverty programs effect (average: 0.138 SD)

Cash transfers effect (average: 0.067 SD)
Overall effect (average: 0.094 SD)

−0.2 −0.1 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7

Treatment Effect (in Standard Deviation Units)

Fig. 4. The impacts of antipoverty programs on mental health. The
estimated treatment effects of RCTs that evaluated antipoverty programs in low-
or middle-income countries on indices of mental health (37–41, 56, 119, 131–149).
Positive treatment effects imply better mental health. “Cash Transfers” refers
to studies of unconditional cash transfers to low-income households, with the
exception of Baird et al. and Kilburn et al., who examined conditional cash
transfer programs (135, 150). “Multifaceted antipoverty programs” refers to
interventions that aimed to lift people out of poverty by providing a range of
elements, typically including asset transfers, skills training, cash support, and access
to savings and health care opportunities. Treatment effects are expressed in SD
units. If multiple follow-up measures were available, this figure shows the final
measure. The outcomes vary across studies and include screening instruments to
detect common mental illnesses [Short General Health Questionnaire (GHQ-12)]
and symptoms of depression [Center for Epidemiologic Studies Depression
Scale (CES-D) and Acholi Psychosocial Assessment (APAI-R)], indices of
psychological well-being (PWB), and a perceived stress scale (PSS). “Intervention
cost” refers to the total cost of implementing the intervention, when this is
available. When implementation costs are unavailable (as with most of the cash
transfer studies), it refers to the total undiscounted value of the transfer. MER,
market exchange rates; PPP, purchasing power parity (which adjusts exchange
rates to reflect the true cost of living). A missing value of years elapsed since
program end means that the transfer was still ongoing when outcomes were
measured. A complete description of the methodology of this analysis and details
on each of the studies is provided in the supplementary materials.

uncertainty and income volatility and jug-
gle what are, in effect, complex financial
portfolios, often without access to formal
insurance (43). Sustained long-run exposure
to stress from managing this volatility may
threaten mental health (44). Consistent with
this hypothesis, a large-scale randomized ex-
periment among low-income individuals in
Oregon found that receiving largely free health
insurance worth $550 to $750 per year re-
duced rates of depression by about a quarter
within a few months (45). This effect did not
appear to be explained by increased mental
health care or changes in physical health. Al-
though the increase in recipients’ effective
income may have played a role, it represented
a much smaller relative increase than the cash
transfer programs described above and yet
generated a similar effect size on depression.
Further suggestive evidence for uncertainty as
a mechanism comes from the small or zero
effect of wealth shocks on mental health in
countries with generous and comprehensive
systems of social insurance, such as Sweden
(34, 35).
Environmental factors
Those living in poverty are generally more
exposed to environmental irritants posed by
pollution, temperature extremes, and chal-
lenging sleep environments (46). Many of
these factors have been linked directly to men-
tal illness. Days with extreme heat see worse

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 4 of 12

RES EARCH | R E V I E W
self-reported mental health and increased
rates of self-harm and suicide (47, 48). Sim-
ilarly, sleep deprivation is widespread among
the urban poor in developing countries (49),
and sleep is thought to be a mechanism af-
fecting mental health (50, 51). Some evi-
dence exists that clinical interventions to
improve sleep reduce depression (49, 52).
Exposure to air pollution is associated with
living in poverty and may influence mental
health through multiple channels, through re-
striction of physical activity or directly owing
to neurotoxicity (53). Changes in air pollution
in China have been associated with changes in
mental health (54). In the United States, ran-
domly selected low-income households that
were paid to move to more affluent neighbor-
hoods saw reductions in depression and anxi-
ety despite little effect on income (55). However,
it is not clear whether environmental factors
or other features of high-income neighbor-
hoods generated this effect.
Physical health
Lower income is robustly associated with
worse physical health (9). Poverty increases
exposure to the environmental factors de-
scribed above and often also implies lower
access to health care, which increases the
burden of acute and chronic health condi-
tions. Worse physical health may affect men-
tal health through various channels. Chronic
pain, worries about health and mortality, the
financial costs of illness, and reduced physi-
cal activity may all worsen mental health. It is
therefore unsurprising that physical ill-health
often co-occurs with depressive and anxiety
disorders (15). However, only limited causal
evidence exists of poverty affecting mental
health through changes in physical health.
Many of the randomized interventions de-
scribed in the previous section had no detec-
table effect on physical health even as they
reduced mental illness (36, 38, 56). However,
changes in physical well-being may manifest
over a longer time frame, which may not be
captured by these short-run studies.
Early-life conditions
Exposure to poverty early in life can threaten
mental health in later years. Such effects can
be generated in utero, by exposing pregnant
women to malnutrition or stress. For example,
the death of a mother’s relative during preg-
nancy (compared with after childbirth) pre-
dicts depression and anxiety among her grown
children later in life (57). Poverty may also
disproportionately expose children to adverse
shocks while their brains are highly plastic.
Such shocks can profoundly affect brain de-
velopment, cognitive ability, and mental health
in adolescence and adulthood (58, 59). Eco-
nomic stresses around the time of birth also
have long-term mental health costs; in Ghana,
Ridley et al., Science 370, eaay0214 (2020) 11 December 2020
a decrease in crop prices by 1 SD at an indi-
vidual’s time of birth was found to increase
incidence of anxiety or depression in adult-
hood by 50%, associating with maternal nutri-
tion, breastfeeding duration, vaccination rates,
and improved adult health (60). These results
imply that programs that provide financial
support for households with pregnant women
or young children may have high long-run
mental health and economic returns.
Trauma, violence, and crime
Living in poverty disproportionately exposes
individuals to crime, including violent offenses
(61). People living in poverty are also more
likely to suffer traumatic events such as the
early deaths of loved ones (62). Likewise, with-
in the household, women and children in poor
households are disproportionately affected
by intimate partner violence (63). The rela-
tionship between poverty and experiencing
violence itself may be causal: Cash transfers to
households reduce intimate partner violence
(64). In turn, both exposure to violence within
the household and exposure to violent crime
in general predict depression and other men-
tal illnesses (65, 66). Causal evidence on the
effect of reductions in crime and violence on
mental illness is needed to shed further light
on this mechanism.
Social status, shame, and isolation
Relative poverty—consumption or income rel-
ative to others in one’s society—may play a
role in the relationship between poverty and
mental illness through the resulting social
status and interpersonal comparisons. In an
interesting natural experiment, Norwegian
tax records were posted online in 2001, mak-
ing citizens’ income easily searchable. Using
survey data from 1985 to 2013, a study showed
that the gap in happiness and life-satisfaction
between the rich and poor within Norway
increased sharply once relative income became
easily visible (67). Although similar causal evi-
dence is lacking for mental illness, it is plau-
sible that diminished social status resulting
from poverty causes or exacerbates depression
and anxiety. Frequent marginalization of peo-
ple living in poverty may also result in social
isolation and loneliness (68), which in turn are
correlated with depression (69).
The causal impact of mental ill-health
on poverty
Mental illness predicts worse labor market
outcomes later in life. After a diagnosis of de-
pression or anxiety, employment rates and
incomes have been estimated to fall by as
much as half, relative to the nondepressed or
nonanxious (70, 71). Beyond such compari-
sons, which may be driven in part by unknown
factors such as physical health, there is little
evidence from natural experiments to link
depression or anxiety to incomes. A study
showed that the approval of lithium for treat-
ment of bipolar disorder reduced the earn-
ings penalty associated with bipolar illness
by a third in Denmark, from 38 to 26%, with
larger effects in the lower half of the earn-
ings distribution (72). Studying similar natural
experiments for depression and anxiety would
be valuable.
There is, however, a substantial body of
experiments that show a causal effect of treat-
ing mental illness on employment. A meta-
analysis that aggregated results across 31 RCTs
in developing countries showed a positive
average effect of various interventions to
treat mental illness on labor supply (73).
Among these interventions, pharmacological
and psychological treatments had similar po-
sitive effects on labor supply (0.1 to 0.15 SD),
and combining both types of treatments had
Downloaded from http://science.sciencemag.org/ on December 10, 2020
even larger effects (0.34 SD). For example, a
cheap and scalable cognitive behavioral the-
rapy (CBT) administered in India reduced de-
pression by 25 percentage points compared
with that in the control group and in turn
increased reported days of work by 2.3 days
per month (Fig. 5 and Box 3). Although these
studies do not directly show that treatment
of mental illness reduces poverty rates, higher
labor supply and earnings naturally reduce
the likelihood of living in poverty. Whether
treatment of mental illness has larger long-
run effects on consumption per unit cost
than the cash transfers described above is
unknown (73).
Mechanisms for mental ill-health
causing poverty
Cognitive function
Like any illness, depression and anxiety may
have economic effects because they directly
reduce individuals’ ability to work. Unlike
most physical conditions, however, depres-
sion and anxiety also directly affect the way
people think. Poverty itself can influence cog-
nitive function by capturing attention and
taxing mental bandwidth (74, 75). Mental
illness may have similar effects, by capturing
attention, causing excessive rumination, and
distorting people’s memories and beliefs about
their abilities (76). Such cognitive impacts
could alter a range of economic decisions
and outcomes, from finding jobs to saving
to education and by exacerbating “behav-
ioral biases” that economists increasingly
recognize as important (77). For example,
depressed individuals might avoid making
active choices and may stick with “default
options,” may have decreased sensitivity to
incentives because of anhedonia, or may have
difficulty choosing among several options.
Understanding the importance of this mech-
anism relative to more “direct” economic ef-
fects through disability or health expenditures
5 of 12
RES EARCH | R E V I E W

Depression (PHQ-9 score > 10) Days unable to work Health costs
1 10 100
| *** | | *** | | * |

$ over 3 months
Days per month
0.8 8 80
Fraction

0.6 6 60

0.4 4 40

0.2 2 20

0 0 0
Control Treatment Control Treatment Control Treatment

Fig. 5. Impacts of behavioral activation on depression, labor supply, and health costs. The mean values for the treatment and control groups of depression
[measured with a Patient Health Questionnaire–9 (PHQ-9) score greater than 10], days unable to work, and health costs (excluding the intervention cost) at 3 months.
Error bars denote ±1 SEM. Asterisks denote a significant difference between treatment and control after adjusting for covariates: *P < 0.1; ***P < 0.01.

on correlations between mental illness and

Box 3. An example of a psychotherapy intervention with positive economic effects.
economic preferences is mixed (83–85).

Downloaded from http://science.sciencemag.org/ on December 10, 2020

Patel et al. conducted an RCT of a brief behavioral activation (BA) therapy program, administered by Labor supply and productivity
nonspecialist counselors in a sample of 495 depressed adults in Goa, India (155). Compared with a
control group that received enhanced usual care, treated patients were >60% more likely to be in Depression and anxiety often affect individ-
remission 3 months later (64 versus 39%), as measured with a PHQ-9 score (Box 1) below 10, and uals in the prime of their economic lives and
maintained these gains after 12 months. can be highly recurrent (86). The depressed
Those patients also reported being able to work 2.3 more days per month on average (P = 0.004) beliefs and distorted preferences described
and reduced health costs, excluding intervention costs, of $20 per month (P = 0.07) (Fig. 5). For above may reduce motivation and labor sup-
comparison, a month’s wages for a low-skilled worker in the study context was around $415. Given an ply. In addition, depression can have a direct
average intervention cost of $66 per patient, in economic terms the intervention was highly cost- effect on productivity, such as through in-
effective and may have paid for itself within a few months. After 12 months, the fall in treated patients’ creased fatigue and lack of concentration.
health costs alone had already significantly outpaced the cost of intervention, although the difference Depressed individuals may therefore work
in days worked was no longer significant (94). fewer and shorter days and produce less per
Other evaluations of inexpensive psychotherapies implemented by nonspecialist counselors in low- hour (87). Depressed workers might also
income settings have found similarly large effects on mental health (122). More evidence on the effect be more easily discouraged during their
of such psychological interventions on broader economic outcomes would be highly valuable. For job search or when facing setbacks at work.
example, future trials could be linked to administrative or standardized survey data on wages, earnings, As described above, substantial causal evi-
and consumption. dence exists that treatment of mental illnesses
increases employment (73). However, there
is little evidence on whether this happens
through higher at-work productivity, greater
is crucial for correctly measuring the eco- risks and thus reduced risk-taking. Such evi- job search intensity, changed beliefs, or other
nomic burden of mental illness and design- dence is consistent with mental illness caus- mechanisms.
ing economic policy for those whose mental ing pessimistic beliefs, pessimistic beliefs
health is compromised. causing mental illness, or both. Causal evi- Stigma
dence on how treatment of depression or Mentally ill individuals contend with sub-
Beliefs anxiety affects beliefs would help disentangle stantial social stigma and negative stereo-
Beliefs about one’s own and others’ abilities, these potential explanations. typing (88). This may result in discrimination
circumstances, and actions are central to eco- in employment (89), which could lower wages
nomic decision-making. Mental illness may Preferences and limit employment opportunities relative
distort such beliefs in various ways. Depres- Mental illness may affect economic prefer- to equally productive mentally healthy work-
sion is associated with more negative beliefs ences, such as the extent to which people are ers. On top of this, those living with a mental
about oneself and the external world (78, 79). willing to defer gratification (time preferen- illness are excluded from disability benefit
Depressed individuals are more likely to re- ces), tolerate risk for higher expected rewards schemes in many low-income countries (90).
member negative stimuli and have trouble (risk preferences), or split rewards between More generally, others’ reluctance to inter-
disengaging from negative information once themselves and others (social preferences). act socially with mentally ill people (88) may
it grabs their attention (76). As such, although For example, depression may diminish a per- exclude them from social networks that pro-
healthy individuals tend to protect overly son’s patience and altruism. Similarly, anxiety vide economic opportunities. Stigma may also
optimistic beliefs about themselves by ignor- disorders may reduce people’s willingness to affect the formation and dissolution of house-
ing negative information (80), correlational take on even modest levels of potentially pro- holds in ways that disadvantage the mentally
evidence suggests that the depressed update fitable risk. Such impacts could in turn change ill (91). Depression and anxiety may come with
their beliefs more pessimistically (81). Anx- a variety of economic behaviors, such as la- a “discount” on the marriage market, causing
iety, meanwhile, is associated with greater bor supply decisions, savings and investment mentally ill individuals to form households
selective attention toward threatening stimuli choices, consumption behavior, and the take- with less well-off partners, increasing the
(82), which could lead to overestimation of up of social programs. The limited evidence chances of living in poverty.

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 6 of 12

RES EARCH | R E V I E W
Health expenditures
Mental illnesses may deepen poverty through
its impacts on health and health expenditures.
In developing countries, people living in pov-
erty usually pay most of their health costs out
of pocket (92). Globally, 150 million people are
estimated to have catastrophic health expen-
ditures each year, which are defined as health
care payments totaling more than 40% of a
household’s nonsubsistence expenditures (93).
Costs associated with treatment of mental
illness rarely account for large shares of in-
dividuals’ budgets because most affected
individuals remain untreated. However, de-
pression and anxiety frequently co-occur with
other health conditions (15), and such comor-
bidity with depression is associated with
substantially higher health expenditures for
a range of health conditions (16, 17). Indeed,
treatment of depression has been found to
reduce overall health care costs (94).
Women’s empowerment
The burden of mental illness falls dispropor-
tionately on women (1). A large-scale (n = 903
participants) RCT that evaluated CBT inter-
vention for depressed pregnant women in
Pakistan found a 17% reduction in depression
rates compared with a control group 7 years
after the intervention (95). Reduced depres-
sion among these women was accompanied by
increased economic empowerment by 0.29 SD
as measured in increased control over house-
hold and personal expenditures. Such im-
pacts may have implications for women’s
consumption and relative poverty within the
household.
Intergenerational effects
Improving a parent’s mental health can bene-
fit the next generation. In the above study in
Pakistan, women who had received the inter-
vention sent their children to better schools
and had more learning materials in their homes
(95). Similarly, other RCTs found that treat-
ment of mothers’ depression improves their
interaction with their children and their child-
ren’s mental health (96). Although little direct
evidence shows that such interventions lead
to improved educational outcomes or earn-
ings, there is reason to believe they may. A
substantial body of work from other con-
texts shows that early-childhood invest-
ments have large effects on children’s income
as adults (97).
Human capital accumulation
The onset of common mental illnesses often
coincides with secondary and tertiary edu-
cation and the early stages of an adult’s work
career (86). Mental illness may therefore cause
long-run economic hardship by reducing school
and college completion rates, worsening early-
career job placements and hindering skill
Ridley et al., Science 370, eaay0214 (2020) 11 December 2020
acquisition (98). This suggests the possibility
of particularly high economic returns from
improving mental health among adolescents
and young adults. Although longitudinal stud-
ies show a substantial correlation between
mental illness among students and subsequent
educational outcomes, there is little experimen-
tal evidence to date that treatment of depres-
sion or anxiety among adolescents leads to
improved educational outcomes (99).
Outlook
Having discussed some of the mechanisms
that influence the relationship between pov-
erty and mental health, we can speculate more
broadly on how the relationship between po-
verty and mental illness may evolve, what this
means for policy, and what directions may be
fruitful for research.
Aggregate economic conditions
Economic growth and other ongoing global
trends are unlikely to improve mental health
by themselves. Higher income causes better
mental health at the individual level, yet on
average, the prevalence of mental illness is
not lower in rich countries. To the contrary,
existing evidence shows a higher prevalence
of common mental illness in richer countries
(Fig. 6) (100). This cross-country difference
cannot be interpreted causally, and concerns
remain about differences in methodology,
diagnosis, or reporting across contexts (101).
However, one way to reconcile the contrast-
ing within-country and cross-country relation-
ships is the possibility that relative, rather
than absolute, poverty is the more relevant
cause of mental illness. Risk factors, includ-
ing inequality and relative poverty, or the
stresses of urban environments, may deterio-
rate rather than improve as whole economies
expand. Within-country inequality has in-
creased in many countries in recent decades,
despite substantial reductions in extreme
poverty and global inequality (102). Com-
placency about mental health among the
poor is therefore not warranted even in the
presence of aggregate economic growth. For
example, the burden of disease attributable
to mental and neurological disorders in India
and China increased by 61 and 28%, respec-
tively, between 1990 and 2013, despite im-
pressive economic growth (103).
Climate change
The more frequent occurrence of extreme heat
because of climate change is anticipated to
exacerbate mental illness directly (104, 105).
The increased frequency of weather-related
disasters, such as floods and hurricanes, poses
a threat to mental health through greater ex-
posure to trauma (106). Climate change also
threatens mental health through its negative
economic consequences, which are likely to
be more pronounced in low-income countries
(104). Extreme temperatures during the ag-
ricultural growing season that damage crops
and thus economic well-being have been re-
ported to increase suicides in agricultural
regions in India (107). Predicted increases in
water scarcity and droughts are also likely to
worsen economic and in turn psychological
well-being. Climate change is also expected to
lead to increased violence and political conflict
over the next century through increased pres-
sure on resources, such as productive land
and, possibly, psychological effects of heat on
aggression (108). This combination of econo-
mic and political consequences of climate
change may increase the flow of refugees and
economic migrants, with concomitant chal-
lenges to mental health (109).
Technological change and globalization
Downloaded from http://science.sciencemag.org/ on December 10, 2020
For many of those living in poverty across the
world, technological change and globalization
offer enormous economic opportunities; how-
ever, both phenomena produce winners and
losers. The costs to losers, especially low-wage
workers in high- and middle-income countries
who lose jobs as a result of changes in patterns
of trade or automation, can be long lasting
and substantial (110), resulting in worse men-
tal health (29, 30). Although most economic
research on these topics focuses on rich coun-
tries, there is an urgent need to understand
the mental-health effects of these economic
changes in poorer countries.
Offering social insurance and welfare, skills
training, and job transition programs, includ-
ing psychotherapies for workers exposed to
the harmful effects of technological change
and globalization, will be important to pro-
tect mental health. One example is the on-
going Building Bridges and Bonds evaluation,
which provides a tailored CBT intervention
to unemployed (formerly incarcerated) men
in the United States, in conjunction with tra-
ditional job-seeking services, in an attempt
to increase employment and wages.
Pandemics
Public-health crises such as the ongoing
COVID-19 pandemic tend to disproportionate-
ly affect those living in poverty (111). They may
worsen mental health on average and partic-
ularly among the poor. Income and employ-
ment losses as a result of morbidity can be
large, which in turn can reduce mental health
through the mechanisms described above. In
addition, the exposure to trauma, increased
worries and uncertainty, and worse physical
health will tend to impair mental health, in
turn reducing income and employment. How-
ever, ongoing progress in reducing the burden
of other diseases that disproportionately af-
fect the poor—such as cholera and diarrhoeal
infections, tuberculosis, malaria, and other
7 of 12
RES EARCH | R E V I E W

Depression Anxiety
8 8

6 6

Point prevalence (%)

Point prevalence (%)

4 4

2 2

Downloaded from http://science.sciencemag.org/ on December 10, 2020

0 0
1 2 4 8 16 32 64 1 2 4 8 16 32 64
Log GDP per capita ($000s PPP) Log GDP per capita ($000s PPP)
Correlation coef. r = 0.582 (p = 0.00) Correlation coef. r = 0.584 (p = 0.00)

Fig. 6. Prevalence of common mental illnesses by country. For each country, the percentage of the population estimated to have a (left) depressive disorder or
(right) anxiety disorder is shown at a given point in time against that country’s log gross domestic product (GDP) per capita. Each scatter point represents one
country. The line shown is an ordinary least squares regression line of country prevalence rates on a constant and log GDP per capita. Prevalence rate data come
from the Global Burden of Disease Study, 2017 (http://ghdx.healthdata.org/gbd-results-tool). GDP per capita data are for 2017, measured in constant 2017
international $, and come from the World Bank’s World Development Indicators dataset (https://databank.worldbank.org/source/world-development-indicators).

insect-borne diseases—will provide a counter-
vailing force likely to improve mental health
among the poor. The COVID-19 pandemic has
the potential to set back progress in reducing
the enormous burden of these diseases.
Social media
The spread of mobile phones and the internet
opens up opportunities for poverty alleviation
(112) and new ways to deliver mental health
care. However, some of these technologies
may pose new threats to mental health. Al-
though more causal evidence is needed, some
studies have found that depression is corre-
lated with internet addiction and with the in-
tensity of use of social media among young
adults and adolescents (113, 114). Recent ran-
domized interventions show that deactivat-
ing social media accounts for 4 weeks led to
0.1 SD reductions in depression and anxiety
scores (115, 116). Increased access to the inter-
net and social media among the global poor,
although presenting enormously valuable be-
nefits, may therefore also pose some threat to
mental health among the poor (117).
Implications for research and policy
Because mental health and poverty are inti-
mately linked, interdisciplinary collaborations
between mental-health researchers and social
scientists studying poverty are essential (Box 4).
Evaluations of economic interventions should
carefully measure impacts on mental health,
and vice versa. An example of such work is
Bhat et al., in which a team of psychiatrists
and economists followed up on psychotherapy
clinical trials in India and deployed the tools of
behavioral economics and psychiatry to study
long-run effects of psychotherapy on mental
health, economic well-being, and decision-
making (118).
Policy tools
There is a strong economic case for investing
in the mental health of people in poverty. A
recent meta-analysis showed that mental health
interventions in low- and middle-income coun-
tries, including psychotherapy and pharma-
cotherapy, had similar or larger effects on
employment than economic interventions
such as cash transfers, job training programs,
and multifaceted antipoverty programs (73).
Yet, they were an order of magnitude less ex-
pensive to deliver. Such treatments could there-
fore be the most cost-effective antipoverty
intervention, at least among the subset of
people who are mentally ill. However, we know
little about how to optimally combine, dose, se-
quence, and target economic and mental health
interventions. For example, combining psycho-
logical and economic interventions may lead to
treatment effects that are greater than the sum
of their parts. In particular, improved mental
health could increase the economic returns of
receiving cash or asset transfers by improving
decision-making and productivity.
Recently, innovative studies have compared
the effects of providing psychotherapy, cash
support, or both among low-income popula-
tions. An RCT measured the effect of 8 weeks
of CBT and/or $200 in cash support to 999
criminally engaged men in Liberia (119). Al-
though the psychotherapy targeted antisocial
behavior rather than mental illness per se, the
study found that the combination of cash trans-
fer and psychotherapy improved an index of
self-regard and mental health by 0.2 SD a year
later (P = 0.024), accompanied by a modest
reduction in depression and psychological dis-
tress (–0.11 SD, P = 0.24). The combined treat-
ment not only reduced antisocial behavior but
also increased patience and self-control, whereas
neither cash nor therapy alone had detectable
effects. However, none of the treatments ap-
parently influenced consumption or income a
year later. More studies along these lines would
be valuable.
Treatment gaps
In poor countries, the fraction of individuals
diagnosed with depression and anxiety who
do not receive treatment often exceeds 90%
(11–14). Such treatment gaps likely result from
a combination of poor supply and low demand
for mental health services.
Increasing supply
Resources for mental health care are extreme-
ly limited in low-income countries (Fig. 7), and

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 8 of 12

RES EARCH | R E V I E W

seek or adhere to treatment (126). People

Box 4. Priorities for future research on poverty and common mental illnesses. often lack mental health literacy and may
not possess basic information about mental
1. Measurement of mental health in economic surveys to estimate health conditions and their risk factors, symp-
The comparative impacts of diverse economic interventions, such as insurance, social safety, toms, and potential treatment options (127).
and employment support, relative to cash transfers Stigma and shame can further depress de-
The longer-run effects of antipoverty programs mand for mental health services. But there
The effects of absolute versus relative poverty are examples of successful community-based
The effect of technological change and globalization on high and low-wage workers programs that increase mental-health liter-
The impact of the looming global economic recession resulting from COVID-19 acy and boost the share of mentally ill indi-
viduals who seek treatment; for example, the
VISHRAM (Vidarbha Stress and Health Pro-
2. Measurement of economic outcomes in intervention studies for depression and anxiety, gramme) initiative, rolled out in 30 villages
which includes in rural India, deployed “change agents”—
Income, labor supply, productivity, and profits from self-employment typically, persons already playing leadership
Economic preferences and beliefs; investment and savings behaviors roles in the villages—to engage the commu-
Household expenditures and consumption, including within-household allocation of resources nity in conversations around mental health
and illness, using a range of contextually ap-
propriate strategies such as group discussions;
3. Evaluations of interventions to reduce stigma and to boost demand for mental health care, the program led to a sixfold increase in help-

Downloaded from http://science.sciencemag.org/ on December 10, 2020

which includes seeking by persons with depression in just
Diverse mental health literacy approaches, from mass-media campaigns to grass-root awareness building 18 months (128). A priority for future work
Subsidies and incentives for seeking and engaging with mental health care should be the evaluation of such programs
The effects of interventions on marginalized and underserved communities at scale, as well as testing approaches that
bundle mental health treatments with other un-
stigmatized services, subsidizing or even reward-
4. Evaluation of technologies to increase the coverage of effective psychotherapies, ing the take-up of treatment, or using remote
which includes technologies such as app-based therapy that
Text, phone, or video delivery are less likely to expose individuals to stigma.
Digital approaches to training, supervision, and quality assurance for frontline providers
Poverty traps
Artifical intelligence bot-based and other guided self-help approaches, adapted to different languages and
cultural contexts A classic idea in development economics is
the existence of a “poverty trap” (129). This is
the idea that below a certain threshold level
5. Evaluation of interventions to interrupt the intergenerational transmission of poverty and of wealth, people are not able to earn and save
mental illness, such as through their way out of poverty. They remain trapped
School mental health–promotion programs and measurement of effects also on educational attainment, in poverty precisely because of the depriva-
labor supply, productivity, and earnings tions associated with poverty and not because
Treatment of parental mental illness and measurement of effects on children’s cognitive and educational of any intrinsic lack of ability. Such poverty
outcomes traps could exist for many reasons. For exam-
ple, the very poor may not be able to afford
enough food to be productive at work. Re-
cent evaluations of multifaceted antipoverty
programs have provided some evidence for
people living in poverty often lack access Reliance on in-person training and super- the existence of such poverty traps (39, 41).
to basic mental health care (120). For ex- vision of the community health workers by However, the underlying mechanisms are un-
ample, some estimates suggest that India experts is a major barrier to scaling these clear. The evidence of a bidirectional causal
has only 3900 psychiatrists for more than approaches. The Empower initiative seeks to relationship between poverty and mental health
1.3 billion people, and 13 psychiatrists serve address these structural barriers through the presented in this Review suggests that mental
Zimbabwe’s 14 million people (121). However, deployment of digital platforms to enable health could be a key mechanism: There could
cost-effective and scalable strategies for treat- frontline workers to learn, deliver, and master be psychological poverty traps. Some of those
ing mental illness in low-resource settings evidence-based brief psychological treatments living in poverty may be ensnared in a vicious
do exist. Evidence from multiple countries for mental health problems, using innovative cycle of poverty and mental illness (5). If this
shows that “psychosocial” treatments such practices such as coach-supported learning and is the case, a one-time economic or psycho-
as manualized talk therapies can be highly peer-supervised quality assurance, all delivered logical intervention of sufficient magnitude
effective at low cost, even when delivered through digital tools (125). Empower will roll may “push” people into a state of sustainably
by nonspecialist community health work- out the brief behavioral activation treatment higher income and better mental health.
ers (122, 123). An even more pared-down— developed and evaluated in India and Nepal
but still effective—approach is the “friendship (Box 3) to community health workers in India Conclusion
benches” of Zimbabwe, in which nonspecialist and the United States from 2021 onward. The causal relationship between poverty and
health workers (popularly known as “commu- mental health we have described could not be
nity grandmothers”) provide problem-solving Stimulating demand more pertinent than in the ongoing pande-
therapy with components of behavior activa- Even in settings with affordable and effective mic, which has already adversely affected both
tion to patients (124). mental health services, many people do not of these outcomes. Given the surge of deaths

Ridley et al., Science 370, eaay0214 (2020) 11 December 2020 9 of 12

RES EARCH | R E V I E W
Psychiatrists
Psychologists
40
Number per 100,000 inhabitants
30
20
10
0 23. J. Das, Q.-T. Do, J. Friedman, D. McKenzie, Mental health
Low Lower middle
Income category
Fig. 7. Availability of mental health workers across countries. The mean numbers of psychiatrists, social
workers, psychologists, and nurses working in the mental health sector (per 100,000 people) are shown
for countries in each of the four income categories used by the World Bank. Data on mental health workers
come from the WHO’s Global Health Observatory (http://apps.who.int/gho/data/node.main.MHHR?lang=en)
and are for the most recent year available (which ranges from 2013 to 2017).
of despair in the United States in the wake of
the Great Recession of 2008 (31), we have
6.
grave concerns about the mental-health impli-
cations of the economic recession confronting
the world. The pandemic has disproportion-
7.
ately affected the poor and may have lasting
adverse impacts on their economic and men-
tal well-being. A massive investment in mental
health was long overdue even before the pan- 8.
demic and has become critically urgent now.
Beyond more money, this is also an important
opportunity to invest wisely in lower-cost in- 9.
novations that provide quality care to low-
income and disadvantaged communities and
10.
to integrate economic interventions with men-
tal health care to reduce historic disparities in
11.
both wealth and mental health.
RE FE RENCES AND N OT ES
1. GBD 2017 Disease and Injury Incidence and Prevalence 12.
Collaborators, Global, regional, and national incidence,
prevalence, and years lived with disability for 354 diseases
and injuries for 195 countries and territories, 1990-2017:
A systematic analysis for the Global Burden of Disease 13.
Study 2017. Lancet 392, 1789–1858 (2018). doi: 10.1016/
S0140-6736(18)32279-7; pmid: 30496104
2. R. Desjarlais, World Mental Health: Problems and Priorities
in Low-income Countries (Oxford Univ. Press, 1995).
3. C. J. L. Murray, A. D. Lopez, Others, “The global burden of 14.
disease: a comprehensive assessment of mortality and
disability from diseases, injuries, and risk factors in 1990 and
projected to 2020: summary” [World Health Organization 15.
(WHO), 1996]; https://apps.who.int/iris/handle/10665/41864.
4. J. Sareen, T. O. Afifi, K. A. McMillan, G. J. G. Asmundson,
Relationship between household income and mental
disorders: Findings from a population-based longitudinal
study. Arch. Gen. Psychiatry 68, 419–427 (2011). 16.
doi: 10.1001/archgenpsychiatry.2011.15; pmid: 21464366
5. C. Lund et al., Poverty and common mental disorders in low and
middle income countries: A systematic review. Soc. Sci. Med.
Ridley et al., Science 370, eaay0214 (2020) 11 December 2020
Mental health social workers
Mental health nurses
Upper middle High
71, 517–528 (2010). doi: 10.1016/j.socscimed.2010.04.027;
pmid: 20621748
V. Iemmi et al., Suicide and poverty in low-income and
middle-income countries: A systematic review. Lancet
Psychiatry 3, 774–783 (2016). doi: 10.1016/S2215-0366(16)
30066-9; pmid: 27475770
L. M. Banks, H. Kuper, S. Polack, Poverty and disability in
low- and middle-income countries: A systematic review.
PLOS ONE 12, e0189996 (2017). doi: 10.1371/
journal.pone.0189996; pmid: 29267388
D. Frasquilho et al., Mental health outcomes in times
of economic recession: A systematic literature review.
BMC Public Health 16, 115 (2016). doi: 10.1186/
s12889-016-2720-y; pmid: 26847554
D. M. Cutler, A. Lleras-Muney, T. Vogl, in The Oxford
Handbook of Health Economics, S. Glied, P. Smith, Eds.
(Oxford Univ. Press, 2011), pp. 124–163.
H. Bleakley, Health, human capital, and development.
Annu. Rev. Econom. 2, 283–310 (2010). doi: 10.1146/
annurev.economics.102308.124436; pmid: 24147187
D. Chisholm et al., Scaling-up treatment of depression
and anxiety: A global return on investment analysis.
Lancet Psychiatry 3, 415–424 (2016). doi: 10.1016/
S2215-0366(16)30024-4; pmid: 27083119
V. Patel, G. Simon, N. Chowdhary, S. Kaaya, R. Araya,
Packages of care for depression in low- and middle-income
countries. PLOS Med. 6, e1000159 (2009). doi: 10.1371/
journal.pmed.1000159; pmid: 19806179
P. S. Wang et al., Use of mental health services for anxiety,
mood, and substance disorders in 17 countries in the
WHO world mental health surveys. Lancet 370,
841–850 (2007). doi: 10.1016/S0140-6736(07)61414-7;
pmid: 17826169
R. Kohn, S. Saxena, I. Levav, B. Saraceno, The treatment gap
in mental health care. Bull. World Health Organ. 82, 858–866
(2004). pmid: 15640922
K. M. Scott et al., Association of mental disorders with
subsequent chronic physical conditions: World mental
health surveys from 17 countries. JAMA Psychiatry 73,
150–158 (2016). doi: 10.1001/jamapsychiatry.2015.2688;
pmid: 26719969
E. P. Vamos, I. Mucsi, A. Keszei, M. S. Kopp, M. Novak,
Comorbid depression is associated with increased healthcare
utilization and lost productivity in persons with diabetes:
A large nationally representative Hungarian population survey.
Psychosom. Med. 71, 501–507 (2009). doi: 10.1097/
PSY.0b013e3181a5a7ad; pmid: 19528291
17. C. Shen, U. Sambamoorthi, G. Rust, Co-occurring mental
illness and health care utilization and expenditures in
adults with obesity and chronic physical illness. Dis. Manag.
11, 153–160 (2008). doi: 10.1089/dis.2007.0012;
pmid: 18564027
18. WHO, “Promoting mental health: concepts, emerging
evidence, practice: a report of the World Health Organization,
Department of Mental Health and Substance Abuse”
(WHO/Victorian Health Promotion Foundation/Univ.
Melbourne, 2005); https://apps.who.int/iris/bitstream/
handle/10665/43286/9241562943-tha.pdf.
19. J. Haushofer, E. Fehr, On the psychology of poverty.
Science 344, 862–867 (2014). doi: 10.1126/science.1232491;
pmid: 24855262
20. R. Layard, J. F. Helliwell, J. D. Sachs, “World Happiness
Report 2012” (Earth Institute, Columbia University, 2012).
21. C. Lund, S. Stansfeld, M. De Silva, Social determinants of
mental health, in Global Mental Health: Principles and Practice
(Oxford Univ. Press, 2014), pp. 116–136.
22. G. Tampubolon, W. Hanandita, Poverty and mental health in
Indonesia. Soc. Sci. Med. 106, 20–27 (2014). doi: 10.1016/
j.socscimed.2014.01.012; pmid: 24524962
Downloaded from http://science.sciencemag.org/ on December 10, 2020
patterns and consequences: Results from survey data in five
developing countries. World Bank Econ. Rev. 23, 31–55
(2009). doi: 10.1093/wber/lhn010
24. J. Corrigall, C. Lund, V. Patel, S. Plagerson, M. K. Funk,
Poverty and mental illness: fact or fiction? A commentary
on Das, Do, Friedman, McKenzie & Scott (65:3, 2007,
467–480). Soc. Sci. Med. 66, 2061–2063, discussion
2064–2066 (2008). doi: 10.1016/j.socscimed.2008.01.005;
pmid: 18325649
25. J. Das, Q.-T. Do, J. Friedman, D. McKenzie, K. Scott,
Revisiting the relationship between mental health and
poverty in developing countries: A response to Corrigall.
Soc. Sci. Med. 66, 2064–2066 (2008). doi: 10.1016/
j.socscimed.2008.01.004
26. S. C. Olesen, P. Butterworth, L. S. Leach, M. Kelaher,
J. Pirkis, Mental health affects future employment as job
loss affects mental health: Findings from a longitudinal
population study. BMC Psychiatry 13, 144 (2013).
doi: 10.1186/1471-244X-13-144; pmid: 23705753
27. M. Alloush, Income, psychological well-being, and the
dynamics of poverty: Evidence from South Africa (2018);
https://ageconsearch.umn.edu/record/274223/files/
Abstracts_18_05_23_17_38_40_27__23_123_4_199_0.pdf.
28. C. Christian, L. Hensel, C. Roth, Income shocks and suicides:
Causal evidence from Indonesia. Rev. Econ. Stat. 101, 1–16
(2019).
29. A. Kuhn, R. Lalive, J. Zweimüller, The public health costs of
job loss. J. Health Econ. 28, 1099–1115 (2009). doi: 10.1016/
j.jhealeco.2009.09.004; pmid: 19833399
30. J. R. Pierce, P. K. Schott, Trade liberalization and mortality:
Evidence from U.S. counties. Am. Econ. Rev. Insights 2,
47–63 (2020). doi: 10.1257/aeri.20180396
31. A. Case, A. Deaton, Deaths of Despair and the Future of
Capitalism (Princeton Univ. Press, 2020).
32. B. Wolfe, J. Jakubowski, R. Haveman, M. Courey, The income
and health effects of tribal casino gaming on American
Indians. Demography 49, 499–524 (2012). doi: 10.1007/
s13524-012-0098-8; pmid: 22427279
33. B. Apouey, A. E. Clark, Winning big but feeling no better?
The effect of lottery prizes on physical and mental health.
Health Econ. 24, 516–538 (2015). doi: 10.1002/hec.3035;
pmid: 24677260
34. D. Cesarini, E. Lindqvist, R. Östling, B. Wallace, Wealth,
health, and child development: Evidence from administrative
data on Swedish lottery players. Q. J. Econ. 131, 687–738
(2016). doi: 10.1093/qje/qjw001
35. E. Lindqvist, R. Östling, D. Cesarini, Long-run effects of
lottery wealth on psychological well-being. Rev. Econ. Stud.
87, 2703–2726 (2020). doi: 10.1093/restud/rdaa006
36. J. Haushofer, J. Shapiro, The short-term impact of
unconditional cash transfers to the poor: Experimental
evidence from Kenya. Q. J. Econ. 131, 1973–2042 (2016).
doi: 10.1093/qje/qjw025; pmid: 33087990
37. J. Haushofer, J. Shapiro, “The long-term impact of
unconditional cash transfers: experimental evidence from
Kenya” (Busara Center for Behavioral Economics, 2018);
https://pdfs.semanticscholar.org/2454/
746163854dd855ac0d93f13f39556232bdf7.pdf.
10 of 12
RES EARCH | R E V I E W
38. A. Banerjee et al., Development economics. A multifaceted
program causes lasting progress for the very poor: Evidence
from six countries. Science 348, 1260799 (2015).
doi: 10.1126/science.1260799; pmid: 25977558
39. O. Bandiera et al., Labor markets and poverty in village
economies. Q. J. Econ. 132, 811–870 (2017). doi: 10.1093/
qje/qjx003
40. G. Bedoya, A. Coville, J. Haushofer, M. R. Isaqzadeh,
J. Shapiro, No Household Left Behind: Afghanistan Targeting
the Ultra Poor Impact Evaluation, policy research working
papers (The World Bank, 2019).
41. A. Banerjee, E. Duflo, R. Chattopadhyay, J. Shapiro, The long
term impacts of a “Graduation” program: Evidence from
West Bengal. Unpublished paper, Massachusetts Institute of
Technology, Cambridge, MA (2016); https://trickleup.org/
wp-content/uploads/2016/12/2016_Banerjee_Long_Term_
Impacts_Bandhan_Full.pdf.
42. A. Banerjee, P. Niehaus, T. Suri, Universal basic income in
the developing world. Annu. Rev. Econ. 11, 959–983 (2019).
doi: 10.1146/annurev-economics-080218-030229
43. D. Collins, J. Morduch, S. Rutherford, O. Ruthven, Portfolios
of the Poor: How the World’s Poor Live on $2 a Day (Princeton
Univ. Press, 2009).
44. S. M. Staufenbiel, B. W. J. H. Penninx, A. T. Spijker,
B. M. Elzinga, E. F. C. van Rossum, Hair cortisol, stress
exposure, and mental health in humans: A systematic review.
Psychoneuroendocrinology 38, 1220–1235 (2013).
doi: 10.1016/j.psyneuen.2012.11.015; pmid: 23253896
45. A. Finkelstein et al., The Oregon health insurance experiment:
Evidence from the first year. Q. J. Econ. 127, 1057–1106
(2012). doi: 10.1093/qje/qjs020; pmid: 23293397
46. E. B. Dean, F. Schilbach, H. Schofield, in The Economics of
Poverty Traps, C. B. Barrett, M. R. Carter, J.-P. Chavas, Eds.
(Univ. Chicago Press, 2018), pp. 57–118.
47. N. Ding, H. Berry, L. O’Brien, The effect of extreme heat on
mental health—Evidence from Australia. Int. J. Epidemiol. 44
(suppl. 1), i64–i64 (2015). doi: 10.1093/ije/dyv097.238
48. M. N. Williams, S. R. Hill, J. Spicer, Do hotter temperatures
increase the incidence of self-harm hospitalisations?
Psychol. Health Med. 21, 226–235 (2016). doi: 10.1080/
13548506.2015.1028945; pmid: 25849087
49. P. Bessone, G. Rao, F. Schilbach, H. Schofield, M. Toma,
The Economic Consequences of Increasing Sleep Among
the Urban Poor, National Bureau of Economic Research (NBER)
working paper no. 26746 (2020); www.nber.org/papers/
w26746.
50. A. G. Harvey, Sleep and circadian functioning: Critical
mechanisms in the mood disorders? Annu. Rev. Clin. Psychol.
7, 297–319 (2011). doi: 10.1146/annurev-clinpsy-032210-
104550; pmid: 21166537
51. S. Grubb, M. Lauritzen, Deep sleep drives brain fluid
oscillations. Science 366, 572–573 (2019). doi: 10.1126/
science.aaz5191; pmid: 31672882
52. R. Manber et al., Cognitive behavioral therapy for insomnia
enhances depression outcome in patients with comorbid
major depressive disorder and insomnia. Sleep 31, 489–495
(2008). doi: 10.1093/sleep/31.4.489; pmid: 18457236
53. Z. Jia et al., Exposure to ambient air particles increases
the risk of mental disorder: Findings from a natural
experiment in Beijing. Int. J. Environ. Res. Public Health 15,
160 (2018). doi: 10.3390/ijerph15010160; pmid: 29351245
54. T. Xue, T. Zhu, Y. Zheng, Q. Zhang, Author correction:
Declines in mental health associated with air pollution and
temperature variability in China. Nat. Commun. 10, 3609
(2019). doi: 10.1038/s41467-019-11660-5; pmid: 31383856
55. J. Ludwig et al., Neighborhood effects on the long-term
well-being of low-income adults. Science 337, 1505–1510
(2012). doi: 10.1126/science.1224648; pmid: 22997331
56. D. Egger, J. Haushofer, E. Miguel, P. Niehaus, M. W. Walker,
General Equilibrium Effects of Cash Transfers: Experimental
Evidence from Kenya. NBER working paper no. 26600
(2019); www.nber.org/papers/w26600.
57. P. Persson, M. Rossin-Slater, Family ruptures, stress, and
the mental health of the next generation. Am. Econ. Rev.
108, 1214–1252 (2018). doi: 10.1257/aer.20141406;
pmid: 30091569
58. K. G. Noble et al., Family income, parental education and
brain structure in children and adolescents. Nat. Neurosci. 18,
773–778 (2015). doi: 10.1038/nn.3983; pmid: 25821911
59. C. Blair, C. C. Raver, Poverty, stress, and brain
development: New directions for prevention and intervention.
Acad. Pediatr. 16 (Suppl), S30–S36 (2016). doi: 10.1016/
j.acap.2016.01.010; pmid: 27044699
Ridley et al., Science 370, eaay0214 (2020) 11 December 2020
60. A. Adhvaryu, J. Fenske, A. Nyshadham, Early life
circumstance and adult mental health. J. Polit. Econ. 127,
1516–1549 (2019). doi: 10.1086/701606
61. P. Sharkey, M. Besbris, M. Friedson, in The Oxford Handbook
of the Social Science of Poverty, D. Brady, L. M. Burton,
Eds. (Oxford Univ. Press, 2016).
62. M. Marmot, Social determinants of health inequalities.
Lancet 365, 1099–1104 (2005). doi: 10.1016/S0140-6736
(05)71146-6; pmid: 15781105
63. C. B. Cunradi, R. Caetano, C. Clark, J. Schafer, Neighborhood
poverty as a predictor of intimate partner violence among
White, Black, and Hispanic couples in the United States:
A multilevel analysis. Ann. Epidemiol. 10, 297–308 (2000).
doi: 10.1016/S1047-2797(00)00052-1; pmid: 10942878
64. J. Haushofer, C. Ringdal, J. P. Shapiro, X. Y. Wang, Income
changes and intimate partner violence: Evidence from
unconditional cash transfers in Kenya. NBER working paper
no. 25627 (2019); www.nber.org/papers/w25627.
65. L. A. Goodman, K. F. Smyth, A. M. Borges, R. Singer, When
crises collide: How intimate partner violence and poverty
intersect to shape women’s mental health and coping?
Trauma Violence Abuse 10, 306–329 (2009). doi: 10.1177/
1524838009339754; pmid: 19776085
66. F. Cornaglia, N. E. Feldman, A. Leigh, Crime and mental
well-being. J. Hum. Resour. 49, 110–140 (2014). doi: 10.3368/
jhr.49.1.110
67. R. Perez-Truglia, The effects of income transparency on well-
being: Evidence from a natural experiment. Am. Econ. Rev.
110, 1019–1054 (2020). doi: 10.1257/aer.20160256
68. R. Walker, G. Bantebya-Kyomuhendo, The Shame of Poverty
(Oxford Univ. Press, 2014).
69. J. T. Cacioppo, M. E. Hughes, L. J. Waite, L. C. Hawkley,
R. A. Thisted, Loneliness as a specific risk factor for
depressive symptoms: Cross-sectional and longitudinal
analyses. Psychol. Aging 21, 140–151 (2006). doi: 10.1037/
0882-7974.21.1.140; pmid: 16594799
70. C. Hakulinen et al., Mental disorders and long-term labour
market outcomes: Nationwide cohort study of 2 055 720
individuals. Acta Psychiatr. Scand. 140, 371–381 (2019).
doi: 10.1111/acps.13067; pmid: 31254386
71. R. Mojtabai et al., Long-term effects of mental disorders on
employment in the National Comorbidity Survey ten-year
follow-up. Soc. Psychiatry Psychiatr. Epidemiol. 50,
1657–1668 (2015). doi: 10.1007/s00127-015-1097-z;
pmid: 26211661
72. B. Biasi, M. S. Dahl, P. Moser, Career effects of mental health.
Working paper (2018); https://papers.ssrn.com/
abstract=2544251.
73. C. Lund, K. Orkin, M. Witte, T. Davies, J. Haushofer, J. Bass,
P. Bolton, S. Murray, L. Murray, W. Tol, G. Thornicroft,
V. Patel, The economic effects of mental health interventions
in low and middle-income countries. Working paper
(2019); https://custom.cvent.com/
4E741122FD8B4A1B97E483EC8BB51CC4/files/Event/
159bd4dc083941a79dd0211437d5d7dc/
d79476623ef7451884092864b1d7116b.pdf.
74. A. Mani, S. Mullainathan, E. Shafir, J. Zhao, Poverty impedes
cognitive function. Science 341, 976–980 (2013).
doi: 10.1126/science.1238041; pmid: 23990553
75. S. Kaur, S. Mullainathan, F. Schilbach, Does financial strain
lower productivity? Working paper (2019); https://
economics.mit.edu/files/16997.
76. I. H. Gotlib, J. Joormann, Cognition and depression: Current
status and future directions. Annu. Rev. Clin. Psychol. 6,
285–312 (2010). doi: 10.1146/annurev.clinpsy.121208.131305;
pmid: 20192795
77. M. Kremer, G. Rao, F. Schilbach, in Handbook of Behavioral
Economics, Foundations and Applications 2, B. D. Bernheim,
S. DellaVigna, D. Laibson, Eds. (North-Holland, 2019),
vol. 2 of Handbooks in Economics, pp. 345–458.
78. A. T. Beck, Depression: Clinical, Experimental, and Theoretical
Aspects (Univ. Pennsylvania Press, 1967).
79. J. de Quidt, J. Haushofer, Depression for economists.
NBER working paper no. 22973 (2016), doi: 10.3386/w22973
80. D. Eil, J. M. Rao, The good news-bad news effect:
Asymmetric processing of objective information about
yourself. Am. Econ. J. Microecon. 3, 114–138 (2011).
doi: 10.1257/mic.3.2.114
81. C. W. Korn, T. Sharot, H. Walter, H. R. Heekeren, R. J. Dolan,
Depression is related to an absence of optimistically
biased belief updating about future life events. Psychol. Med.
44, 579–592 (2014). doi: 10.1017/S0033291713001074;
pmid: 23672737
82. T. Armstrong, B. O. Olatunji, Eye tracking of attention in the
affective disorders: A meta-analytic review and synthesis.
Clin. Psychol. Rev. 32, 704–723 (2012). doi: 10.1016/
j.cpr.2012.09.004; pmid: 23059623
83. Y. M. Bayer, Z. Shtudiner, O. Suhorukov, N. Grisaru, Time
and risk preferences, and consumption decisions of patients
with clinical depression. J. Behav. Exp. Econ. 78, 138–145
(2019). doi: 10.1016/j.socec.2019.01.003
84. D. A. Cobb-Clark, S. Dahmann, N. Kettlewell, “Depression,
risk preferences and risk-taking behavior,” research paper
12285 (Institute for the Study of Labor, 2019); https://
papers.ssrn.com/abstract=3390275.
85. J. S. Rounds, J. G. Beck, D. M. Grant, Is the delay
discounting paradigm useful in understanding social
anxiety? Behav. Res. Ther. 45, 729–735 (2007).
doi: 10.1016/j.brat.2006.06.007; pmid: 16890909
86. R. C. Kessler et al., Age of onset of mental disorders:
A review of recent literature. Curr. Opin. Psychiatry 20,
359–364 (2007). doi: 10.1097/YCO.0b013e32816ebc8c;
pmid: 17551351
87. S. Mall et al., Days out of role due to mental and physical
illness in the South African stress and health study. Soc.
Psychiatry Psychiatr. Epidemiol. 50, 461–468 (2015).
doi: 10.1007/s00127-014-0941-x; pmid: 25096982
88. B. A. Pescosolido, T. R. Medina, J. K. Martin, J. S. Long,
The “backbone” of stigma: Identifying the global core
Downloaded from http://science.sciencemag.org/ on December 10, 2020
of public prejudice associated with mental illness.
Am. J. Public Health 103, 853–860 (2013). doi: 10.2105/
AJPH.2012.301147; pmid: 23488508
89. J. Sharac, P. McCrone, S. Clement, G. Thornicroft,
The economic impact of mental health stigma and
discrimination: A systematic review. Epidemiol. Psichiatr.
Soc. 19, 223–232 (2010). doi: 10.1017/
S1121189X00001159; pmid: 21261218
90. S. Saxena, P. Sharan, M. Garrido, B. Saraceno, World Health
Organization’s Mental Health Atlas 2005: Implications for
policy development. World Psychiatry 5, 179–184 (2006).
pmid: 17139355
91. C. Lauber, W. Rössler, Stigma towards people with mental
illness in developing countries in Asia. Int. Rev. Psychiatry
19, 157–178 (2007). doi: 10.1080/09540260701278903;
pmid: 17464793
92. A. V. Banerjee, E. Duflo, The economic lives of the poor.
J. Econ. Perspect. 21, 141–167 (2007). doi: 10.1257/
jep.21.1.141; pmid: 19212450
93. K. Xu et al., Protecting households from catastrophic health
spending. Health Aff. (Millwood) 26, 972–983 (2007).
doi: 10.1377/hlthaff.26.4.972; pmid: 17630440
94. B. Weobong et al., Sustained effectiveness and cost-
effectiveness of the Healthy Activity Programme, a brief
psychological treatment for depression delivered by lay
counsellors in primary care: 12-month follow-up of a
randomised controlled trial. PLOS Med. 14, e1002385 (2017).
doi: 10.1371/journal.pmed.1002385; pmid: 28898283
95. V. Baranov, S. R. Bhalotra, P. Biroli, J. Maselko, Maternal
depression, women’s empowerment, and parental
investment: Evidence from a large randomized control trial
(2017); https://papers.ssrn.com/abstract=3081421.
96. P. Cuijpers, E. Weitz, E. Karyotaki, J. Garber, G. Andersson,
The effects of psychological treatment of maternal
depression on children and parental functioning: A meta-
analysis. Eur. Child Adolesc. Psychiatry 24, 237–245 (2015).
doi: 10.1007/s00787-014-0660-6; pmid: 25522839
97. J. J. Heckman, S. Mosso, The economics of human
development and social mobility. Annu. Rev. Econom. 6,
689–733 (2014). doi: 10.1146/annurev-economics-080213-
040753; pmid: 25346785
98. G. C. Patton et al., Our future: A Lancet commission on
adolescent health and wellbeing. Lancet 387, 2423–2478
(2016). doi: 10.1016/S0140-6736(16)00579-1; pmid: 27174304
99. D. Prinz, M. Chernew, D. Cutler, A. Frakt, “Health and economic
activity over the lifecycle: Literature review,” NBER working paper
series 24865 (NBER, 2018), doi: 10.3386/w24865
100. M. L. A. Dückers, L. Reifels, D. P. De Beurs, C. R. Brewin,
The vulnerability paradox in global mental health and its
applicability to suicide. Br. J. Psychiatry 215, 588–593
(2019). doi: 10.1192/bjp.2019.41
101. P. Brhlikova, A. M. Pollock, R. Manners, Global burden of
disease estimates of depression—How reliable is the
epidemiological evidence? J. R. Soc. Med. 104, 25–34 (2011).
doi: 10.1258/jrsm.2010.100080; pmid: 21205775
102. B. Milanovic, Global Inequality: A New Approach for the Age of
Globalization (Harvard Univ. Press, 2016).
11 of 12
RES EARCH | R E V I E W
103. F. J. Charlson, A. J. Baxter, H. G. Cheng, R. Shidhaye,
H. A. Whiteford, The burden of mental, neurological, and
substance use disorders in China and India: A systematic 124.
analysis of community representative epidemiological
studies. Lancet 388, 376–389 (2016). doi: 10.1016/
S0140-6736(16)30590-6; pmid: 27209143
104. V. Masson-Delmotte, Global Warming of 1.5 OC: An IPCC 125.
Special Report on the Impacts of Global Warming of 1.5° C
Above Pre-industrial Levels and Related Global Greenhouse
Gas Emission Pathways, in the Context of Strengthening the 126.
Global Response to the Threat of Climate Change, Sustainable
Development, and Efforts to Eradicate Poverty (World
Meteorological Organization, 2018).
105. P. M. Trang, J. Rocklöv, K. B. Giang, G. Kullgren, M. Nilsson, 127.
Heatwaves and hospital admissions for mental disorders in
northern Vietnam. PLOS ONE 11, e0155609 (2016).
doi: 10.1371/journal.pone.0155609; pmid: 27195473 128.
106. H. L. Berry, T. D. Waite, K. B. G. Dear, A. G. Capon, V. Murray,
The case for systems thinking about climate change and
mental health. Nat. Clim. Chang. 8, 282–290 (2018).
doi: 10.1038/s41558-018-0102-4
107. T. A. Carleton, Crop-damaging temperatures increase
suicide rates in India. Proc. Natl. Acad. Sci. U.S.A. 114, 129.
8746–8751 (2017). doi: 10.1073/pnas.1701354114;
pmid: 28760983
108. M. Burke, S. M. Hsiang, E. Miguel, Climate and conflict. 130.
Annu. Rev. Econ. 7, 577–617 (2015). doi: 10.1146/
annurev-economics-080614-115430
109. C. McMichael, J. Barnett, A. J. McMichael, An ill wind?
Climate change, migration, and health. Environ. Health
Perspect. 120, 646–654 (2012). doi: 10.1289/ehp.1104375; 131.
pmid: 22266739
110. D. H. Autor, D. Dorn, G. H. Hanson, The China shock: Learning
from labor-market adjustment to large changes in trade.
Annu. Rev. Econ. 8, 205–240 (2016). doi: 10.1146/
annurev-economics-080315-015041 132.
111. A. Adams-Prassl, T. Boneva, M. Golin, C. Rauh, “Inequality in
the Impact of the Coronavirus Shock: Evidence from Real
Time Surveys,” Cambridge-INET working paper no. 2020/18
(2020); https://papers.ssrn.com/sol3/papers.cfm?
abstract_id=3594297. 133.
112. T. Suri, W. Jack, The long-run poverty and gender impacts of
mobile money. Science 354, 1288–1292 (2016). doi: 10.1126/
science.aah5309; pmid: 27940873
113. J. H. Ha et al., Depression and Internet addiction in 134.
adolescents. Psychopathology 40, 424–430 (2007).
doi: 10.1159/000107426; pmid: 17709972
114. L. Y. Lin et al., Association between social media use
and depression among U.S. young adults. Depress. Anxiety
33, 323–331 (2016). doi: 10.1002/da.22466; 135.
pmid: 26783723
115. H. Allcott, L. Braghieri, S. Eichmeyer, M. Gentzkow, The
welfare effects of social media. Am. Econ. Rev. 110, 629–676 136.
(2020). doi: 10.1257/aer.20190658
116. R. Mosquera, M. Odunowo, T. McNamara, X. Guo, R. Petrie,
The economic effects of Facebook. Exp. Econ. 10.2139/
ssrn.3312462 (2019).
117. World Bank Group, World Development Report 2016: Digital 137.
Dividends (World Bank Publications, 2016).
118. B. Bhat, J. Haushofer, J. de Quidt, V. Patel, F. Schilbach,
G. Rao, P.-L. Vautrey, The economics consequences of
depression: Evidence from Goa. AEA RCT Registry (2019), 138.
doi: 10.1257/rct.3823-1.0
119. C. Blattman, J. C. Jamison, M. Sheridan, Reducing crime and
violence: Experimental evidence from cognitive behavioral
therapy in Liberia. Am. Econ. Rev. 107, 1165–1206 (2017).
doi: 10.1257/aer.20150503; pmid: 29553237
120. M. Thirunavukarasu, P. Thirunavukarasu, Training and
national deficit of psychiatrists in India—A critical analysis. 139.
Indian J. Psychiatry 52 (suppl. 1), S83–S88 (2010).
doi: 10.4103/0019-5545.69218; pmid: 21836723
121. WHO, “Mental health atlas 2017” (WHO, 2017);available at
www.who.int/mental_health/evidence/atlas/ 140.
mental_health_atlas_2017/en.
122. D. R. Singla et al., Psychological treatments for the world:
Lessons from low- and middle-income countries. Annu. Rev.
Clin. Psychol. 13, 149–181 (2017). doi: 10.1146/ 141.
annurev-clinpsy-032816-045217; pmid: 28482687
123. V. Patel, N. Chowdhary, A. Rahman, H. Verdeli, Improving
access to psychological treatments: Lessons from developing
Ridley et al., Science 370, eaay0214 (2020) 11 December 2020
countries. Behav. Res. Ther. 49, 523–528 (2011).
doi: 10.1016/j.brat.2011.06.012; pmid: 21788012
D. Chibanda et al., Effect of a primary care–based psychological
intervention on symptoms of common mental disorders in
Zimbabwe: A randomized clinical trial. JAMA 316, 2618–2626
(2016). doi: 10.1001/jama.2016.19102; pmid: 28027368
R. Waters, Community workers lend human connection to
COVID-19 response. Health Aff. (Millwood) 39, 1112–1117
(2020). doi: 10.1377/hlthaff.2020.00836; pmid: 32634347
V. Patel et al., The magnitude of and health system
responses to the mental health treatment gap in adults in
India and China. Lancet 388, 3074–3084 (2016).
doi: 10.1016/S0140-6736(16)00160-4; pmid: 27209149
A. F. Jorm, Mental health literacy. Public knowledge and
beliefs about mental disorders. Br. J. Psychiatry 177, 396–401
(2000). doi: 10.1192/bjp.177.5.396; pmid: 11059991
R. Shidhaye et al., The effect of VISHRAM, a grass-roots
community-based mental health programme, on the
treatment gap for depression in rural communities in India:
A population-based study. Lancet Psychiatry 4, 128–135
(2017). doi: 10.1016/S2215-0366(16)30424-2;
pmid: 28063879
P. Dasgupta, D. Ray, Inequality as a determinant of
malnutrition and unemployment: Theory. Econ. J. (Lond.) 96,
1011 (1986). doi: 10.2307/2233171
B. A. Arvind et al., T. M. Shibukumar, NMHS Collaborators Group,
Prevalence and socioeconomic impact of depressive disorders
in India: Multisite population-based cross-sectional study. BMJ
Open 9, e027250 (2019). doi: 10.1136/bmjopen-2018-027250;
pmid: 31253618
E. P. Green, C. Blattman, J. Jamison, J. Annan, Does poverty
alleviation decrease depression symptoms in post-conflict
settings? A cluster-randomized trial of microenterprise
assistance in Northern Uganda. Glob. Ment. Health (Camb.) 3,
e7 (2016). doi: 10.1017/gmh.2015.28; pmid: 28596876
C. Blattman, S. Dercon, S. Franklin, Impacts of industrial
and entrepreneurial jobs on youth: 5-year experimental
evidence on factory job offers and cash grants in “Ethiopia,”
NBER working paper no. 25788 (2019); www.nber.org/
papers/w25788.
C. Blattman, N. Fiala, S. Martinez, “The Long Term Impacts of
Grants on Poverty: 9-year Evidence From Uganda’s Youth
Opportunities Program,” NBER working paper no. 24999
(2018); www.nber.org/papers/w24999.
L. Hjelm, S. Handa, J. de Hoop, T. Palermo; Zambia CGP
and MCP Evaluation Teams, Poverty and perceived
stress: Evidence from two unconditional cash transfer
programs in Zambia. Soc. Sci. Med. 177, 110–117 (2017).
doi: 10.1016/j.socscimed.2017.01.023; pmid: 28167339
S. Baird, B. de Hoop, B. Özler, Özler, Income shocks and
adolescent mental health. J. Hum. Resour. 48, 370–403
(2013). doi: 10.3368/jhr.48.2.370
K. Kilburn, H. Thirumurthy, C. T. Halpern, A. Pettifor, S. Handa,
Effects of a large-scale unconditional cash transfer
program on mental health outcomes of young people in
Kenya. J. Adolesc. Health 58, 223–229 (2016). doi: 10.1016/
j.jadohealth.2015.09.023; pmid: 26576822
G. Angeles et al., Government of Malawi’s unconditional
cash transfer improves youth mental health. Soc. Sci. Med.
225, 108–119 (2019). doi: 10.1016/j.socscimed.2019.01.037;
pmid: 30826585
J. Haushofer, R. Mudida, J. Shapiro, “The comparative
impact of cash transfers and psychotherapy on
psychological and economic well-being,” working paper
(2020); www.researchgate.net/profile/Robert_Mudida/
publication/340435415_The_Comparative_Impact_of_
Cash_Transfers_and_Psychotherapy_on_Psychological_
and_Economic_Well-being.
J. Haushofer, M. Chemin, C. Jang, J. Abraham, Economic and
psychological effects of health insurance and cash transfers:
Evidence from a randomized experiment in Kenya. J. Dev. Econ.
144, 102416 (2020). doi: 10.1016/j.jdeveco.2019.102416
E. Edmonds, C. Theoharides, The short term impact of a
productive asset transfer in families with child labor:
Experimental evidence from the Philippines. J. Dev. Econ.
146, 102486 (2020). doi: 10.1016/j.jdeveco.2020.102486
S. Janzen, N. Magnan, S. Sharma, W. Thompson, Short-term
impacts of a pay-it-forward livestock transfer and training
program in Nepal. AEA Pap. Proc. 108, 422–425 (2018).
doi: 10.1257/pandp.20181120
142. N. Glass, N. A. Perrin, A. Kohli, J. Campbell, M. M. Remy,
Randomised controlled trial of a livestock productive asset
transfer programme to improve economic and health
outcomes and reduce intimate partner violence in a
postconflict setting. BMJ Glob. Health 2, e000165 (2017).
doi: 10.1136/bmjgh-2016-000165; pmid: 28589002
143. S. Baird, C. McIntosh, B. Özler, When the money runs out: Do
cash transfers have sustained effects on human capital
accumulation? 30 March 2019 (socialprotection.org, 2019);
https://socialprotection.org/sites/default/files/
publications_files/Baird%20McIntosh%20Ozler_When%
20the%20Money%20Runs%
20Out_COMBINED_Revision_FINAL.pdf.
144. J. Shapiro, The impact of recipient choice on aid
effectiveness. World Dev. 116, 137–149 (2019). doi: 10.1016/
j.worlddev.2018.10.010; pmid: 30944504
145. A. Aguilar, Cash and conditionalities’ effects on early
childhood: medium-term effects on physical and cognitive
development (2012); www.aguilaresteva.com/wp-content/
uploads/2012/09/Progresa_ECD_paper.pdf.
146. K. Macours, R. Vakis, Changing Households’ Investments and
Aspirations Through Social Interactions: Evidence From a
Randomized Transfer Program (The World Bank, 2009).
147. L. C. H. Fernald, M. Hidrobo, Effect of Ecuador’s
cash transfer program (Bono de Desarrollo Humano)
Downloaded from http://science.sciencemag.org/ on December 10, 2020
on child development in infants and toddlers: A
randomized effectiveness trial. Soc. Sci. Med. 72,
1437–1446 (2011). doi: 10.1016/j.socscimed.2011.03.005;
pmid: 21531060
148. L. Ismayilova et al., Improving mental health among ultra-poor
children: Two-year outcomes of a cluster-randomized trial in
Burkina Faso. Soc. Sci. Med. 208, 180–189 (2018). doi: 10.1016/
j.socscimed.2018.04.022; pmid: 29743136
149. R. Heath, M. Hidrobo, S. Roy, Cash transfers, polygamy, and
intimate partner violence: Experimental evidence from Mali.
J. Dev. Econ. 143, 102410 (2020). doi: 10.1016/
j.jdeveco.2019.102410
150. K. Kilburn et al., Cash transfers, young women’s economic
well-being, and HIV risk: Evidence from HPTN 068. AIDS
Behav. 23, 1178–1194 (2019). doi: 10.1007/s10461-018-2329-5;
pmid: 30415429
151. V. Patel et al., Detecting common mental disorders in
primary care in India: A comparison of five screening
questionnaires. Psychol. Med. 38, 221–228 (2008).
doi: 10.1017/S0033291707002334; pmid: 18047768
152. K. Kroenke, R. L. Spitzer, The PHQ-9: A new depression
diagnostic and severity measure. Psychiatr. Ann. 32, 509–515
(2002). doi: 10.3928/0048-5713-20020901-06
153. R. L. Spitzer, K. Kroenke, J. B. W. Williams, B. Löwe, A brief
measure for assessing generalized anxiety disorder: The
GAD-7. Arch. Intern. Med. 166, 1092–1097 (2006).
doi: 10.1001/archinte.166.10.1092; pmid: 16717171
154. G.-C. Ali, G. Ryan, M. J. De Silva, Validated screening tools for
common mental disorders in low and middle income
countries: A systematic review. PLOS ONE 11, e0156939
(2016). doi: 10.1371/journal.pone.0156939; pmid: 27310297
155. V. Patel et al., The Healthy Activity Program (HAP), a lay
counsellor-delivered brief psychological treatment for severe
depression, in primary care in India: A randomised controlled
trial. Lancet 389, 176–185 (2017). doi: 10.1016/S0140-6736
(16)31589-6; pmid: 27988143
AC KNOWLED GME NTS
We thank V. Baranov, T. Cruz Vital, J. Das, E. Gallagher, J. Haushofer,
A. Karing, J. Li, C. Lund, M. Mani, K. Orkin, J. Pan, and K. Rao,
for helpful comments and suggestions. We thank six anonymous
referees for detailed comments and helpful suggestions. We
thank C. Roth and L. Hensel for kindly providing us the data
needed to construct Fig. 3. Funding: None declared. Author
contributions: All authors contributed equally to this work.
Competing interests: The authors have no competing interests
or financial conflicts of interest.
SUPPLEMENTARY MATERIALS
science.sciencemag.org/content/370/6522/eaay0214/suppl/DC1
Supplementary Text
Figs. S1 to S3
References (156–164)
10.1126/science.aay0214
12 of 12
Poverty, depression, and anxiety: Causal evidence and mechanisms
Matthew Ridley, Gautam Rao, Frank Schilbach and Vikram Patel

Science 370 (6522), eaay0214.

DOI: 10.1126/science.aay0214

Taxing mental health

Mental equilibrium is essential for an economically productive life in both industrialized and developing countries.
Accumulating evidence shows that mental ill-health and poverty tend to be traveling partners, but which is the cause?
Ridley et al. reviewed the literature on natural and controlled economic experiments involving individuals living in poverty.

Downloaded from http://science.sciencemag.org/ on December 10, 2020

The authors sought to resolve the mechanisms whereby poverty triggers mental illness and how mental illness
compounds poverty. Their results reveal the benefits of cash support and of low-cost therapeutic interventions for those
suffering from mental illness under poverty.
Science, this issue p. eaay0214

ARTICLE TOOLS http://science.sciencemag.org/content/370/6522/eaay0214

SUPPLEMENTARY http://science.sciencemag.org/content/suppl/2020/12/09/370.6522.eaay0214.DC1
MATERIALS

REFERENCES This article cites 120 articles, 13 of which you can access for free
http://science.sciencemag.org/content/370/6522/eaay0214#BIBL

PERMISSIONS http://www.sciencemag.org/help/reprints-and-permissions

Use of this article is subject to the Terms of Service

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published by the American Association for the Advancement of
Science, 1200 New York Avenue NW, Washington, DC 20005. The title Science is a registered trademark of AAAS.
Copyright © 2020 The Authors, some rights reserved; exclusive licensee American Association for the Advancement of
Science. No claim to original U.S. Government Works